Large-scale cancer genomics projects such as The Cancer Genome Atlas (TCGA), International Cancer Genome Consortium (ICGC), Therapeutically Applicable Research to Generate Effective Treatments (TARGET), and the Pediatric Cancer Genome Project (PCGP) are producing a wealth of high throughput sequence data from a large number of cancer samples and their matched normals. These data hold great promise for understanding the genetic basis of cancer and also for the identification of germline susceptibility variants in cancer. Major advancements have been made to systematically catalog somatic variations in cancer genomes from these data sets. However, identifying and interpreting germline changes using data from these studies remains a significant challenge. The primary difficulty stems from 1) the lack of computational pipelines/tools to utilize tumor and normal sequencing data for simultaneous detection of somatic, germline, and LOH events at the nucleotide and chromosomal levels and 2) the lack of uniform bioinformatics analysis strategies for identifying and prioritizing deleterious candidate germline variants responsible for susceptibility. We will develop a computational pipeline for the identification and interpretation f germline alterations in cancer including single nucleotide variants, insertions and deletions (indels), copy number variations, and structural variants. This pipeline will be initially used to systematically analyze whole genome, exome, and RNA-sequencing data from over 5,000 cancer cases already generated by several major efforts and individual research groups and additional cases that will be made publicly available in the next several years. In silico predicte deleterious germline variants from these data will be used for statistical association analysis across groups stratified by age and cancer type to identify novel germline susceptibility variants, genes, and pathways involved in different cancer types. We will further investigate the potential interaction between germline susceptibility variants and somatic mutational landscape. Finally, both pipeline and results from this project will be made publically available, facilitating the analysis and interpretation by the research community of the ever- growing large-scale cancer sequencing data to better discover and understand germline susceptibility variants.